Field
[0001] The present disclosure relates to multidrop bus networks, and in particular concerns
an apparatus and associated method of collision avoidance for a node of a multidrop
bus network.
Background
[0002] Collisions on the bus of a multidrop bus network can cause a decrease in bus efficiency
and throughput. The apparatus and associated method described herein may address this
issue.
Summary
[0003] According to a first aspect of the present disclosure, there is provided a reconciliation
module for a node of a multidrop bus network, the node comprising a MAC module and
a PHY module, the reconciliation module comprising circuitry configured to:
assert a sense signal until the beginning of each transmit opportunity of the node;
de-assert the sense signal at that moment to cause the MAC module to send available
data to the reconciliation module to start transmission of the data on the multidrop
bus network via the PHY module at the respective transmit opportunity; and
re-assert the sense signal on receipt of the data from the MAC module or time-out
of the respective transmit opportunity.
[0004] The MAC module may be configured to delay sending of the data by a wait time after
de-assertion of the sense signal, and the reconciliation module may be configured
to pass the data received from the MAC module to the PHY module for transmission on
the multidrop bus network without transmitting a filler sequence during the wait time
of the MAC module.
[0005] In one or more embodiments, each transmit opportunity may have a duration which is
equal to or greater than the wait time of the MAC module.
[0006] In one or more embodiments, the wait time of the MAC module may be equal to an inter-packet-gap
length.
[0007] In one or more embodiments, in the event the MAC module sends data to the reconciliation
module that cannot be transmitted on the multidrop bus network starting at a transmit
opportunity, the reconciliation module may be configured to:
assert a collision signal on receipt of the data from the MAC module to stop the MAC
module from continuing to send the data; and
control the sense signal to cause the MAC module to resend the data for transmission
starting at the next transmit opportunity.
[0008] In one or more embodiments, the data sent by the MAC module may comprise a plurality
of data frames within the same sequence for transmission starting at a transmit opportunity,
and the reconciliation module may be configured to insert an active-idle sequence
between adjacent data frames to prevent the multidrop bus network from becoming idle
before the plurality of data frames have been transmitted.
[0009] In one or more embodiments, the reconciliation module may be bufferless.
[0010] According to a second aspect of the present disclosure, there is provided a node
of a multidrop bus network, the node comprising the reconciliation module of the first
aspect, the MAC module and the PHY module.
[0011] In one or more embodiments, the multidrop bus network may be an Ethernet network,
the MAC module may be an Ethernet CSMA/CD MAC layer and the PHY module may be a 10BASE-T1S
PHY layer.
[0012] According to a third aspect of the present disclosure, there is provided a multidrop
bus network comprising a plurality of the nodes of the second aspect.
[0013] According to a fourth aspect of the present disclosure, there is provided a method
of collision avoidance for a node of a multidrop bus network, the node comprising
a reconciliation module, a MAC module and a PHY module, the method comprising:
asserting, by the reconciliation module, a sense signal until the beginning of each
transmit opportunity of the node;
de-asserting, by the reconciliation module, the sense signal at that moment to cause
the MAC module to send available data to the reconciliation module to start transmission
of the data on the multidrop bus network via the PHY module at the respective transmit
opportunity; and
re-asserting, by the reconciliation module, the sense signal on receipt of the data
from the MAC module or time-out of the respective transmit opportunity.
[0014] According to a fifth aspect of the present disclosure, there is provided a computer
program comprising computer code configured to control the reconciliation module of
the first aspect, control the node of the second aspect, control the multidrop bus
network of the third aspect, or perform the method of the fourth aspect.
[0015] While the disclosure is amenable to various modifications and alternative forms,
specifics thereof have been shown by way of example in the drawings and will be described
in detail. It should be understood, however, that other embodiments, beyond the particular
embodiments described, are possible as well. All modifications, equivalents, and alternative
embodiments falling within the spirit and scope of the appended claims are covered
as well.
[0016] The above discussion is not intended to represent every example embodiment or every
implementation within the scope of the current or future Claim sets. The figures and
Detailed Description that follow also exemplify various example embodiments. Various
example embodiments may be more completely understood in consideration of the following
Detailed Description in connection with the accompanying Drawings.
Brief Description of the Drawings
[0017] One or more embodiments will now be described by way of example only with reference
to the accompanying drawings in which:
Figure 1 shows the IEEE-standard media-independent interface (MII) between MAC and
PHY modules;
Figure 2 shows a reconciliation module between the MAC and PHY modules to avoid bus
collisions, wherein both the MAC and PHY modules communicate with the reconciliation
module using a MII;
Figure 3 shows a reconciliation module with data buffer between the MAC and PHY modules;
Figure 4 shows transmission attempts by the MAC module of the node of Figure 3 on
a fully-loaded bus;
Figure 5 shows transmission attempts by the MAC module of the node of Figure 3 in
which a transmit opportunity has timed out;
Figure 6 shows transmission attempts by the MAC module of the node of Figure 3 in
which consecutive transmit opportunities have timed out;
Figure 7 shows a reconciliation module without data buffer between the MAC and PHY
modules;
Figure 8 shows transmission attempts by the MAC module of the node of Figure 7 on
a fully-loaded bus;
Figure 9 shows transmission attempts by the MAC module of the node of Figure 7 in
which a transmit opportunity has timed out;
Figure 10 shows transmission attempts by the MAC module of the node of Figure 7 in
which consecutive transmit opportunities have timed out;
Figure 11 shows a method of collision avoidance for a node of a multidrop bus network;
and
Figure 12 shows a computer-readable medium comprising a computer program configured
to control, perform or enable the method of Figure 11.
Detailed Description
Ethernet standards
[0018] While much of the focus in recent Ethernet development has centred on high data rates,
not every application requires speeds of up to 400 Gbps. For some applications, including
Internet of Things (IoT), industrial and automotive, 10 Mbps is sufficient for certain
functions. Factors like cost, weight, distance and the space required for cables are
more important for these use cases.
[0019] Recognizing these evolving requirements, IEEE began work in early 2017 to define
IEEE 802.3cg, a standard for single-pair Ethernet that supports 10 Mbps. The goals
of IEEE 802.3cg were to define a point-to-point and a multidrop short-distance standard
with a maximum length of 25 meters, and a long-distance point-to-point standard that
supports distances up to 1,000 meters. The resulting IEEE 802.3cg specification includes
two link-layer standards: 10BASE-T1S and 10BASE-T1L.
[0020] The 10BASE-T1S short-range standard is primarily targeted at automotive and industrial
applications. Multiple nodes on the network can share a cable in half-duplex shared-medium
mode (multidrop mode) using the standard Ethernet Carrier-Sense Multiple Access with
Collision Detection (CSMA/CD) access method or operate using PHY-Level Collision Avoidance
(PLCA). The cable might be an unshielded twisted pair (UTP) that may have multiple
nodes with their medium dependent interfaces (MDIs) attached thereto. As such, 10BASE-T1S
may also be referred to as Multidrop Single Pair Ethernet.
[0021] The 10BASE-T1L long-range option is designed for IoT and industrial control applications.
The 1,000-meter range is sufficient for use in large factories or warehouses, and
10 Mbps is sufficient for gathering data from sensors and to monitor and control many
types of industrial machinery. It shares the advantages of the short network variant:
MAC-level compatibility with multipair Ethernet and lower cost, weight and required
space.
[0022] The following description relates to the 10BASE-T1S standard but may be also applicable
to other (including future) networking standards and is therefore not necessarily
limited to Ethernet or 10BASE-T1S.
Technical problem
[0023] Figure 1 illustrates the carrier-sense multiple access with collision detection (CSMA/CD)
media access control (MAC) protocol. CSMA/CD uses carrier-sensing (CRS) to defer transmissions
until no other nodes on the network are transmitting. When a collision is detected
by the PHY module/layer of a node (CRS=1), it stops transmitting that frame, asserts
a collision signal (COL=1) to the MAC module/layer of the node, and waits until the
bus is silent before de-asserting the sense signal (CRS=0). The MAC module keeps transmitting
for a predefined (jam) time to ensure that all nodes on the bus will observe the collision
and then pauses for a predefined time interval before trying to resend the frame to
the PHY module for re-transmission on the bus. The first re-transmission attempt occurs
one IPG time after the sense signal (CRS) is de-asserted. If that would result in
another collision, however, then the MAC module introduces a random variable back-off
time from the second re-transmission attempt onwards to avoid indefinitely repeated
collisions.
[0024] A problem with CSMA/CD is that multiple nodes may attempt to transmit data on the
bus simultaneously when the bus is silent. This leads to collisions, corrupted data,
time-outs and loss of bus capacity, which in turn can result in severe degradation
of effective bus throughput or even a complete collapse as the total bus load gets
closer to (e.g. around 70% of) the maximum bus capacity.
[0025] Figure 2 shows an example of a collision avoidance mechanism developed to improve
bus capacity. This mechanism is known as physical layer collision avoidance (PLCA)
and uses a reconciliation module/layer connected between the MAC and PHY modules with
respective IEEE-standard media-independent interfaces (MIIs). Each MII may comprise
the same set of pins: transmit data (TXD[3:0]), transmit enable (TXEN), transmit error
(TXER), receive data (RXD[3:0]), received data valid (RXDV), receive error (RXER),
collision detect (COL) and carrier sense (CRS), and furthermore two clock signals
(TXCLK and RXCLK, not shown) as defined in IEEE 802.3 Clause 22.
[0026] PLCA uses a round-robin scheme in which the nodes are numbered and given the opportunity
to transmit one at a time starting at a dedicated time slot (or transmit opportunity
TO) in a cyclic manner based on their numbering. One node is assigned as the head
node (or PLCA coordinator) and starts each cycle with a beacon. If no data is transmitted
by a node at its transmit opportunity, the transmit opportunity is passed to the next
node. The reconciliation module knows when the transmit opportunity of its node occurs
based on the node number and manages the MAC module using collision (COL-M) and sense
(CRS-M) signal manipulations to prevent the MAC module from sending data that cannot
be transmitted by the PHY module. In this way, the reconciliation module only passes
data to the PHY module when there is a transmit opportunity. The combination of the
round-robin scheme with the reconciliation module therefore avoids collisions on the
bus. Since there should be no collision events for the PHY module to report, the COL-P
pin becomes functionally redundant (and is hence greyed out in Figure 3) but may still
be useful for detecting true bus problems.
[0027] Figure 3 shows an example of a PLCA mechanism involving a reconciliation module with
data buffer (e.g. a first in first out buffer). The data buffer addresses a scenario
in which the reconciliation module starts to receive data from the MAC module shortly
before a transmit opportunity of the node. In this scenario, the data has been sent
from the MAC module too early to be transmitted immediately but too late for the reconciliation
module to signal a collision and have the data re-transmitted in time for the upcoming
transmit opportunity. The data buffer therefore enables the reconciliation module
to temporarily store data sent from the MAC module so that it can be passed to the
PHY module for transmission at the upcoming transmit opportunity.
[0028] Figure 4 shows transmission attempts (MAC-TXDATA) by the MAC module of a node with
the collision avoidance mechanism of Figure 3 on a fully-loaded bus, wherein frames
are timely available in the MAC modules of the respective nodes to utilise every transmit
opportunity. In this figure, four nodes are transmitting data on the bus with each
node transmitting a single data frame at each of its respective transmit opportunities.
The numbering of the nodes is provided for illustrative purposes and is non-limiting
(e.g. the head node may be numbered "0" in practice). The first transmission attempts
of node 2 for data frames 1-3 are respectively denoted F1-F3 and the second (re-transmission
attempts are respectively denoted F1'-F3'.
[0029] The reconciliation module receives a detect signal (CRS-P) from the PHY module indicative
of transmission activity detected on the bus. The detect signal substantially follows
the bus schedule with a slight delay depending on the responsiveness of the PHY module
to detect and report the bus activity. The reconciliation module is configured to
de-assert the sense signal (CRS-M) each time the detect signal (CRS-P) is de-asserted,
unless there has been an earlier transmission attempt by the MAC module for the same
frame as described below.
[0030] The MAC module is configured to send data to the reconciliation module whenever it
has a frame pending for transmission and the sense signal (CRS-M) has been de-asserted
by the reconciliation module. To ensure a minimum gap between frames for most Ethernet
PHY types, however, the MAC module will delay sending the data to the reconciliation
module by a wait time after de-assertion of the sense signal (CRS-M). In this example
shown, the wait time of the MAC module is equal to an inter-packet-gap (IPG) length
(e.g. 12 bytes). The reconciliation module then re-asserts the sense signal (CRS-M)
again on receipt of the data from the MAC module or on detecting activity on the bus
via the detect signal (CRS-P). In the examples shown herein, de-assertion of the sense
signal (CRS-M) is directly correlated with de-assertion of the detect signal (CRS-P),
but re-assertion of the sense signal (CRS-M) is slightly delayed relative to re-assertion
of the detect signal (CRS-P). This is intended to help ensure that the MAC module
sends the pending data after the wait time even if the transmit opportunities are
shorter than 1 IPG but may not be necessary. As such, re-assertion of the sense signal
(CRS-M) could also be directly correlated with re-assertion of the detect signal (CRS-P).
[0031] While the sense signal (CRS-M) is de-asserted, the data from the MAC module is buffered
in the reconciliation module provided the buffer does not overflow. If a transmission
attempt by the MAC module cannot be sufficiently aligned with a transmit opportunity
for that node using the buffer, the reconciliation module is configured to assert
a collision signal (COL-M) to stop the MAC module from continuing to send the data.
This happens by default when another node starts transmitting on the bus and the detect
signal (CRS-P) is asserted. The reconciliation module then de-asserts the collision
signal (COL-M). The collision signal (COL-M) need only be asserted long enough to
be recognised by the MAC module, and should ideally be as short as possible (e.g.
one clock cycle) to avoid any loss of time until recovery.
[0032] Following assertion of the collision signal (COL-M) by the reconciliation module,
the sense signal (CRS-M) is not immediately de-asserted. This is for two reasons.
Firstly, the MAC module requires a recovery time (CR) following assertion of the collision
signal (COL-M) before it can resend the data as part of the upcoming transmission
attempt. Secondly, it is more efficient for the bus to try and ensure that the re-transmission
attempt is synchronised with the upcoming transmit opportunity to avoid any further
delays in transmitting the data frame. To achieve this, the sense signal (CRS-M) is
kept asserted until the data can be directly transmitted without buffering at the
upcoming transmit opportunity and then de-asserted to enable said transmission after
the recovery time of the MAC module.
[0033] The data frame is then passed by the reconciliation module to the PHY module for
transmission (PHY-TXDATA) starting at the upcoming transmit opportunity. It should
be mentioned that transmit opportunities can be relatively short (e.g. shorter than
the IPG length) and typically do not provide enough time for a node to transmit complete
frames. Rather, they provide a time slot or window for the node to claim the bus before
the opportunity times out and the next node has the opportunity. A node can claim
the bus during a transmit opportunity by starting to transmit a data frame or a filler
sequence (SYNC symbols). For example, the reconciliation module may be configured
to transmit SYNC symbols on the bus via the PHY module during the wait time of the
MAC module. Once the bus has been claimed by the node, it can then proceed to transmit
any frame size or even multiple data frames if configured to allow this.
[0034] In the example of Figure 4, the upcoming transmit opportunity for frame 1 of node
2 is the next transmit opportunity on the bus schedule following the (virtual) collision.
For frames 2 and 3 of node 2, on the other hand, there are intervening transmit opportunities
for other nodes on the bus schedule. Since the reconciliation module keeps the sense
signal (CRS-M) asserted until the frames can be re-transmitted at the respective upcoming
transmit opportunities, the de-assertions of the sense signal (CRS-M) which would
otherwise follow the detect signal (CRS-P) from the PHY module are skipped.
[0035] Figure 5 shows further transmission attempts by the MAC module of a node with the
collision avoidance mechanism of Figure 3. Unlike the example of Figure 4 in which
every transmit opportunity on the bus is used (i.e. a fully-loaded bus), one of the
transmit opportunities for node 2 has timed out. This is because frame 3 of node 2
only becomes available in the MAC module sometime during the transmission of frame
3 of node 1. Since the MAC module delays sending data to the reconciliation module
by a wait time of one IPG after de-assertion of the sense signal (CRS-M), there is
insufficient time to start transmitting frame 3 in the upcoming transmit opportunity
for that node if the transmit opportunity ends within roughly 1 IPG length after de-assertion
of the sense signal (CRS-M). As such, the reconciliation module asserts the collision
(COL-M) and sense (CRS-M) signals and keeps the sense signal (CRS-M) asserted until
the next transmit opportunity of the node is detected, at which point it is de-asserted
to enable transmission of the frame on the bus. It is important to note here that,
although the reconciliation module can temporarily store a limited amount of data
in the buffer, the buffer is flushed if any other node transmits data between the
transmit opportunities of this node (otherwise a large buffer to store complete frames
would be required in the reconciliation module). For this reason, frame 3 could not
be stored in the buffer in preparation for the next transmit opportunity of node 2
(hence why the collision signal (COL-M) was asserted and the first transmission attempt
aborted).
[0036] Figure 6 shows further transmission attempts by the MAC module of a node with the
collision avoidance mechanism of Figure 3. In this example, consecutive transmit opportunities
for nodes 3, 4 and 1 have each timed out such that node 2 is the only node transmitting
on the bus for a duration. This is referred to herein as a "boomerang transmit opportunity"
for node 2 where the transmit opportunity for this node returns without any intervening
transmissions by other nodes.
[0037] As shown in the figure, the MAC module attempts to transmit frame 1 an IPG length
after the sense signal (CRS-M) is de-asserted following transmission of the first
frame by node 1. Since this is too late for transmission at the upcoming transmit
opportunity of node 2, the reconciliation module temporarily stores frame 1 in the
buffer until the next transmit opportunity arrives, at which point it is transmitted
on the bus by the PHY module. This is only possible because there are no intervening
transmissions by other nodes, otherwise the buffer would have been flushed and a collision
signal (COL-M) asserted (as per the example in Figure 5). The result is the transmission
of frames 1-3 by node 2 being delayed by the buffer until the respective transmit
opportunities of that node without the need for any re-transmission attempts by the
MAC module. In contrast, collision signals (COL-M) are asserted for frames 4 and 5
because they are received from the MAC module while there are intervening transmissions
by other nodes. These frames are therefore re-transmitted at the next transmit opportunity
similar to the previous examples.
[0038] Although the PLCA mechanism successfully avoids collisions on the bus by blocking
transmission outside of the scheduled transmit opportunities for the node, the use
of a data buffer in the reconciliation module complicates the system and loses its
benefit when the MAC module tries to send more data - before it can be transmitted
on the bus - than the buffer can temporarily store. Furthermore, the time-period during
which it is too early to transmit the data from the MAC module and yet too late for
re-transmission after a virtual collision is relatively short in relation to the total
bus cycle time. Also, the head node may transmit a beacon at the beginning of every
bus cycle as mentioned previously. In this scenario, unless the head node is the only
transmitting node, or the head node sends a short fixed-length beacon but no data
frame, the buffer could potentially be flushed by transmission of the beacon thus
preventing the boomerang transmit opportunity illustrated in Figure 6. In view of
the above, it can be difficult to justify the additional cost and complexity associated
with the data buffer.
Embodiments of the present disclosure
[0039] Figure 7 shows an example of a collision avoidance mechanism involving a reconciliation
module (or layer) which avoids collisions on the bus without requiring a data buffer.
As in the previous examples, the reconciliation module is configured to receive data
sent from the MAC module (or layer) for transmission on the multidrop bus network
via the PHY module (or layer) at a transmit opportunity of the node. The MAC module
is configured to send data to the reconciliation module whenever it has a frame pending
for transmission, the sense signal (CRS-M) has been de-asserted by the reconciliation
module, and the wait time has passed following de-assertion of the sense signal (CRS-M).
[0040] Also, in the event that the data is received from the MAC module while not being
able to transmit it via the PHY module, the reconciliation module is configured to
assert a collision signal (COL-M), keep the sense signal (CRS-M) asserted until the
data can be transmitted at the upcoming transmit opportunity, and then de-assert the
sense signal (CRS-M) at the upcoming transmit opportunity to enable said transmission.
[0041] Unlike the earlier collision avoidance mechanism, however, the reconciliation module
does not de-assert the sense signal (CRS-M) each time the detect signal (CRS-P) is
de-asserted. Instead, the reconciliation module is configured to de-assert the sense
signal (CRS-M) only at the beginning of each transmit opportunity of the node. In
particular, the sense signal (CRS-M) is de-asserted at that moment to cause the MAC
module to send available data to the reconciliation module to start transmission of
the data at the respective transmit opportunity, and then re-asserted on receipt of
the data from the MAC module.
[0042] This approach ensures that data which is already pending in the MAC module is passed
to the reconciliation module at just the right time to enable it to be transmitted
at the upcoming transmit opportunity of the node as a first transmission attempt.
Furthermore, it avoids the need for the reconciliation module to buffer the data,
and also avoids the need for collision signals to be asserted for those data frames
that are available (pending) in advance of the upcoming transmit opportunity. For
this scheme to work, the separation of frames on the bus should be equal to or larger
than 1 IPG. The benefit is simplicity, but the penalty is additional quiet time on
the bus between frames.
[0043] Figure 8 shows transmission attempts (MAC-TXDATA) by the MAC module of a node (node
2 in this example), and the associated data transmission (PHY-TXDATA) by the PHY module
of the node, with the collision avoidance mechanism of Figure 7 on a fully-loaded
bus for comparison with Figure 4. Although the bus schedule is shown with a fixed
bus cycle time and fixed frame lengths for the nodes, this is merely an example and
is not necessary. In this figure, the wait time of each MAC module is equal to one
IPG length.
[0044] The duration of a transmit opportunity is a variable to be set by a system integrator
and depends on several parameters. In practice, it would likely be around 0.5 IPG
lengths for the collision avoidance mechanism of Figure 3. With the collision avoidance
mechanism of Figure 7, however, each transmit opportunity must have a duration of
equal to or greater than the wait time of the MAC module, which is typically one IPG
length. This increases the chance that frames will be received in sufficient time
for transmission on the first attempt. Furthermore, it keeps the minimum distance
between frames on the bus at or above 1 IPG length, like for many other Ethernet PHY
modules/layers. For first transmission attempts, the reconciliation module has no
knowledge of the pending frame. As such, it cannot claim the bus by sending SYNC symbols
(i.e. a filler sequence) to the PHY module during the wait time as it could do for
re-transmission attempts when it knows for sure that a frame is coming from the MAC
module. In addition, the longer transmit opportunity eliminates the need for the reconciliation
module to claim the bus anyway before a frame gets transmitted by the MAC, as the
MAC wait time before receiving a frame from the MAC can be accommodated.
[0045] As shown in Figure 8, frames 1-3 of node 2 are available in the MAC module at the
start of the respective transmit opportunity. As such, they are passed to the reconciliation
module once the wait time has expired and transmitted on the first attempt. Also,
even though the transmit opportunities and silent spacing between transmissions of
different nodes may be longer than the example of Figure 4, the efficiency of the
bus (PHY-TXDATA) is unaffected because filler sequences are not required to claim
the bus.
[0046] Figure 9 shows transmission attempts by the MAC module of node 2 with the collision
avoidance mechanism of Figure 7 for comparison with Figure 5. Like the example of
Figure 5, the transmit opportunity for frame 3 of node 2 times out because this frame
arrives too late for transmission at the transmit opportunity (i.e. frame 3 is not
pending/available in the MAC module until well into the transmit opportunity period).
As a result, the reconciliation module is forced to assert a collision signal (COL-M)
on receipt of the data from the MAC module to stop it from continuing to send the
data, and then control the sense signal (CRS-M) to cause the MAC module to resend
the data for transmission at the next transmit opportunity (similar to before). This
demonstrates that even with longer transmit opportunities, there can still be occasions
where the data frames are received from the MAC module too late to be transmitted
at the first transmit opportunity and need to be delayed until the next (second) transmit
opportunity. Again, there is no impact on the bus efficiency relative to Figure 5.
[0047] Figure 10 shows further transmission attempts by the MAC module of node 2 with the
collision avoidance mechanism of Figure 7 for comparison with Figure 6. Again, consecutive
transmit opportunities for nodes 3, 4 and 1 have each timed out such that node 2 is
the only node transmitting on the bus for a duration (i.e. boomerang transmit opportunities).
This time, however, instead of the reconciliation module buffering frames 1-3 which
are received from the MAC module too early for immediate transmission, the asserted
sense signal (CRS-M) prevents the MAC module from sending the frames until the transmit
opportunity arrives. As such, the reconciliation module receives these frames immediately
after the wait time of the MAC module, which are then passed to the PHY module for
transmission on the first attempt at their respective transmit opportunity.
[0048] Regarding frames 4 and 5, keeping the sense signal (CRS-M) asserted between consecutive
transmissions of the same node means that the MAC module cannot pass these frames
to the reconciliation module while other nodes are transmitting. For this reason,
the assertion of collision signals and re-transmission of frames 4 and 5 as in Figure
6 are avoided.
[0049] As illustrated with the above examples, the collision avoidance mechanism of Figure
7 provides a simpler approach to preventing collisions on the bus than existing systems
by removing the need for the reconciliation module to buffer data or transmit SYNC
symbols. Furthermore, increasing the length of the transmit opportunities for the
node(s) increases the number of frames that are transmitted at the first transmit
opportunity thereby reducing the number of aborted frames (using COL-M) and re-transmission
attempts.
[0050] Although the illustrated examples show each transmit opportunity limited to the transmission
of a single data frame, it is also possible for a node to transmit a plurality of
data frames within a single sequence starting at one transmit opportunity. To allow
for this, the reconciliation module may be configured to insert an active-idle sequence
between adjacent data frames of the sequence to prevent the bus from becoming idle
before the plurality of data frames have been transmitted.
[0051] Each node of the multidrop bus network described herein may comprise a MAC module,
a PHY module and a reconciliation module. The multidrop bus network may be an Ethernet
network, the MAC module may be an Ethernet CSMA/CD MAC layer and the PHY module may
be a 10BASE-T1S PHY layer. Furthermore, given that the reconciliation module described
in relation to Figure 7 does not require a data buffer, the reconciliation module
may be bufferless. Even in "bufferless" examples without a data buffer for storing
frames until the next transmit opportunity, the reconciliation module may still comprise
a small buffer of a few bits in size to facilitate parsing of data between layers/modules
of the Ethernet protocol.
[0052] Although the reconciliation module (or layer) has been described herein with reference
to the MAC and PHY modules (or layers), it could be formed independently of the other
modules and incorporated into the node during a subsequent modular assembly process.
In some cases, the reconciliation module may even be retrofit to an existing node.
Alternatively, the reconciliation module may be integrated with the other modules
of the node at the time of manufacture (e.g. as different parts/portions of the same
chip). One or more (or each) module may comprise circuitry including at least one
processor and at least one memory including computer program code. The at least one
memory and computer program code may be configured to, with the at least one processor,
cause the module to perform the associated functionality described herein.
[0053] Figure 11 illustrates schematically the main steps of a method of collision avoidance
for a node of a multidrop bus network, the node comprising a reconciliation module,
a MAC module and a PHY module. As shown, the method comprises:
asserting, by the reconciliation module, a sense signal until the beginning of each
transmit opportunity of the node;
de-asserting, by the reconciliation module, the sense signal at that moment to cause
the MAC module to send available data to the reconciliation module to start transmission
of the data on the multidrop bus network via the PHY module at the respective transmit
opportunity; and
re-asserting, by the reconciliation module, the sense signal on receipt of the data
from the MAC module or time-out of the respective transmit opportunity.
[0054] Figure 12 illustrates schematically a computer/processor readable medium providing
a computer program according to one example. The computer program may comprise computer
code configured to perform, control or enable one or more of the method steps of Figure
11 using a reconciliation module described herein. In this example, the computer/processor
readable medium is a disc such as a digital versatile disc (DVD) or a compact disc
(CD). In other embodiments, the computer/processor readable medium may be any medium
that has been programmed in such a way as to carry out an inventive function. The
computer/processor readable medium may be a removable memory device such as a memory
stick or memory card (SD, mini SD, micro SD or nano SD) or an embedded or external
non-volatile memory (e.g. flash or MRAM).
[0055] The instructions and/or flowchart steps in the above figures can be executed in any
order, unless a specific order is explicitly stated. Also, those skilled in the art
will recognize that while one example set of instructions/method has been discussed,
the material in this specification can be combined in a variety of ways to yield other
examples as well, and are to be understood within a context provided by this detailed
description.
[0056] In some example embodiments the set of instructions/method steps described above
are implemented as functional and software instructions embodied as a set of executable
instructions which are effected on a computer or machine which is programmed with
and controlled by said executable instructions. Such instructions are loaded for execution
on a processor (such as one or more CPUs). The term processor includes microprocessors,
microcontrollers, processor modules or subsystems (including one or more microprocessors
or microcontrollers), or other control or computing devices. A processor can refer
to a single component or to plural components.
[0057] In other examples, the set of instructions/methods illustrated herein and data and
instructions associated therewith are stored in respective storage devices, which
are implemented as one or more non-transient machine or computer-readable or computer-usable
storage media or mediums. Such computer-readable or computer usable storage medium
or media is (are) considered to be part of an article (or article of manufacture).
An article or article of manufacture can refer to any manufactured single component
or multiple components. The non-transient machine or computer usable media or mediums
as defined herein excludes signals, but such media or mediums may be capable of receiving
and processing information from signals and/or other transient mediums.
[0058] Example embodiments of the material discussed in this specification can be implemented
in whole or in part through network, computer, or data based devices and/or services.
These may include cloud, internet, intranet, mobile, desktop, processor, look-up table,
microcontroller, consumer equipment, infrastructure, or other enabling devices and
services.
[0059] In one example, one or more instructions or steps discussed herein are automated.
The terms automated or automatically (and like variations thereof) mean controlled
operation of an apparatus, system, and/or process using computers and/or mechanical/electrical
devices without the necessity of human intervention, observation, effort and/or decision.
[0060] It will be appreciated that any components said to be coupled may be coupled or connected
either directly or indirectly. In the case of indirect coupling, additional components
may be located between the two components that are said to be coupled.
[0061] In this specification, example embodiments have been presented in terms of a selected
set of details. However, a person of ordinary skill in the art would understand that
many other example embodiments may be practiced which include a different selected
set of these details. It is intended that the following claims cover all possible
example embodiments.
1. A reconciliation module for a node of a multidrop bus network, the node comprising
a MAC module and a PHY module, the reconciliation module comprising circuitry configured
to:
assert a sense signal until the beginning of each transmit opportunity of the node;
de-assert the sense signal at that moment to cause the MAC module to send available
data to the reconciliation module to start transmission of the data on the multidrop
bus network via the PHY module at the respective transmit opportunity; and
re-assert the sense signal on receipt of the data from the MAC module or time-out
of the respective transmit opportunity.
2. The reconciliation module of claim 1, wherein the MAC module is configured to delay
sending of the data by a wait time after de-assertion of the sense signal, and wherein
the reconciliation module is configured to pass the data received from the MAC module
to the PHY module for transmission on the multidrop bus network without transmitting
a filler sequence during the wait time of the MAC module.
3. The reconciliation module of claim 2, wherein each transmit opportunity has a duration
which is equal to or greater than the wait time of the MAC module.
4. The reconciliation module of claim 2 or 3, wherein the wait time of the MAC module
is equal to an inter-packet-gap length.
5. The reconciliation module of any preceding claim, wherein, in the event the MAC module
sends data to the reconciliation module that cannot be transmitted on the multidrop
bus network starting at a transmit opportunity, the reconciliation module is configured
to:
assert a collision signal on receipt of the data from the MAC module to stop the MAC
module from continuing to send the data; and
control the sense signal to cause the MAC module to resend the data for transmission
starting at the next transmit opportunity.
6. The reconciliation module of any preceding claim, wherein the data sent by the MAC
module comprises a plurality of data frames within the same sequence for transmission
starting at a transmit opportunity, and wherein the reconciliation module is configured
to insert an active-idle sequence between adjacent data frames to prevent the multidrop
bus network from becoming idle before the plurality of data frames have been transmitted.
7. The reconciliation module of any preceding claim, wherein the reconciliation module
is bufferless.
8. A node of a multidrop bus network, the node comprising the reconciliation module of
any preceding claim, the MAC module and the PHY module.
9. The node of claim 8, wherein the multidrop bus network is an Ethernet network, the
MAC module is an Ethernet CSMA/CD MAC layer and the PHY module is a 10BASE-T1S PHY
layer.
10. A multidrop bus network comprising a plurality of nodes as defined in claim 8 or 9.
11. A method of collision avoidance for a node of a multidrop bus network, the node comprising
a reconciliation module, a MAC module and a PHY module, the method comprising:
asserting, by the reconciliation module, a sense signal until the beginning of each
transmit opportunity of the node;
de-asserting, by the reconciliation module, the sense signal at that moment to cause
the MAC module to send available data to the reconciliation module to start transmission
of the data on the multidrop bus network via the PHY module at the respective transmit
opportunity; and
re-asserting, by the reconciliation module, the sense signal on receipt of the data
from the MAC module or time-out of the respective transmit opportunity.
12. A computer program comprising computer code configured to perform the method of claim
11.